The development of functional magnetic resonance imaging (fMRI) has greatly expanded our ability to examine neural processing while subjects perform a variety of perceptual, cognitive, or motor tasks. fMRI permits the simultaneous examination of functional activity in many different brain regions and provides a better understanding of how multiple regions work in concert to produce particular behaviors. The next necessary step in the evolution of this approach is to deactivate a specific region in the brain, while a subject is being scanned, to determine the contribution of this site to the activity of its functional network -- by examining the activity of other network components in the absence of this region. To accomplish this goal, we propose to develop a reversible deactivation approach that is fMRI-compatible. The development of this technique will be based on a well-established cryoloop technique in which a cooling device is chronically implanted in behaving animals. We propose a four-phase development program. In phase 1 (development), we will develop a non-metallic cooling device capable of deactivating specific regions of the cerebral cortex. In phase 2 (testing), we will test the device in the behaving cat by cooling regions of primary visual cortex (V1) while the animal performs a visual orienting task. Once we establish device reliability, and can produce behavioral deficits, we will anesthetize the cats and examine the function of the device in the fMRI to confirm compatibility. In phase 3 (implementation), we will chronically implant the device over a region of V1 in non-human primates that have been trained to make visually guided saccades. If V1 is successfully deactivated, saccades to targets within the scotoma should be abolished. The monkeys will be tested while performing the task in the magnet and images will be collected before, during, and after cooling deactivation. We will also examine the distant deactivation consequences on area MTN5 and the superior colliculus while the monkeys view expanding and contracting visual flow during fixation. Finally, in phase 4 (evaluation), we will administer radiolabelled 2-deoxyglucose during cooling so that we may compare the autoradiograms with the deactivation images collected in the magnet. In summary, we expect to fully develop and test a device to reversibly deactivate sites in the brain during behavioral testing in the magnet. Such a development should have wide appeal and will greatly advance our ability to dissect functional circuits with fMRI. [unreadable] [unreadable]